Does early improvement predict endpoint response in patients with generalized anxiety disorder (GAD) treated with pregabalin or venlafaxine XR

Abstract Many patients with generalized anxiety disorder (GAD) only respond to pharmacological treatment after a delay of some weeks, and approximately 35% of patients do not respond. Therefore, early identification of potential responders may have important implications for clinical decision-making. In order to identify early improvement criteria that optimally predict eventual response during short-term treatment of GAD with pregabalin or venlafaxine XR, data were pooled from four double-blind, placebo-controlled GAD treatment studies. A range of measures were analyzed using logistic regression models and receiver operator characteristic (ROC) curve analysis, to predict endpoint response. Results showed that early improvement (≥ 20% reduction from baseline score) on the Hamilton Anxiety Scale (HAM-A) was associated with a high probability of achieving an endpoint response at Weeks 1 and 2 among patients treated with pregabalin (~ 67%), and at Week 2 with venlafaxine XR (60%). A Clinical Global Impression – Improvement (CGI-I) score ≤ 3 at Week 2 was a reliable predictor of achieving endpoint response for pregabalin and venlafaxine XR (odds ratio [OR], 5.33 and 2.47, respectively) with high sensitivity (pregabalin, 0.91; venlafaxine XR, 0.86) and relatively low specificity (pregabalin, 0.33; venlafaxine XR, 0.29), indicating a high true positive rate, but relatively low true negative rate. These findings indicate that improvement by Week 2 on the single item CGI may be a simple and reliable way to predict treatment response with pregabalin or venlafaxine XR in patients with GAD, but a less reliable way to predict non-responders.